Measuring the Internet Skills of Gen Z Students in Higher Education: Validation of the Internet Skills Scale in University Settings

Internet technologies have infiltrated higher education institutions around the world. At the same time, the latest generation of students, the so-called Generation Z (Gen Z), are entering higher education. Gen Z is the first generation born in an Internet-connected world, and digital devices are a seamless part of its life. As a result, Gen Z students have already been engaged with informal digital learning via internet-based technologies outside of formalized education settings. However, previous research has shown that their engagement with these technologies is limited and might not sufficiently cover the knowledge and skills needed to perform internet activities effectively in higher education. Additionally, their familiarity with digital devices and tools varies. Consequently, there is a need for higher education institutions to close the skills gap by applying assessment processes that will assist them in forming policies and training resources for undergraduate students. To achieve the above, research efforts need to focus on developing theoretically informed and valid instruments that measure internet skills. This study has contributed to the validation of a self-assessment questionnaire, the Internet Skills Scale, that can be used in university settings. The questionnaire measures five types of internet skills: operational, information-navigation, social, creative, and critical. The results presented herein provide directions for future research in the field.


1.Introduction
Today's university students belong to the so-called Generation Z.This generation was born in the 1990s and has experienced a globally connected world where the Internet has always been available (Seemiller, & Grace, 2016).As a result, this generation, who is also reffered to as "digital natives", has already used various digital technologies from a very young age, and the Internet is a seamless part of its life (Seemiller & Grace, 2016;Mohr & Mohr, 2017).Although young people's use of digital technologies takes place outside formal education settings, such technologies are also an essential part of their academic lives (Bullen, Morgan, & Qayyum, 2011;Ng, 2012;Ting, 2015;Brooks, 2016;Alexander, Becker, Cummins, & Giesinger, 2017).Despite that, research indicates that informal experiences with digital technologies do not necessarily suggest that undergraduate students can use such technologies for academic purposes (Kennedy & Fox, 2013;Šorgo, Bartol, Dolničar, & Boh Podgornik, 2017).One explanation for this may be that such learning experiences might not sufficiently cover all the skills an individual must acquire to exploit the full potential of digital technologies (van Laar, van Deursen, van Dijk & de Haan 2017).Concequently, many students enter higher education institutions with limited digital skills (Sewlyn, 2009;Ng, 2012;Ståhl, 2017).Additionally, the level of students' digital skills and their familiarity with digital devices and tools varies (Helsper & Eynon, 2010;Hagittai, 2010;Corrin, Bennett, & Lockyer, 2010).
It is widely recognized by higher education institutions and organizations that digital skills are considered an essential component of the human capital, and highly competent users are thus in a better position to be benefited (van Laar, van Deursen, van Dijk, & de Haan 2017).According to forecasts of the European Commission, the future workforce is expected to possess a basic level of digital skills for at least 90% of the professions in the future (European Commission, 2017).In this regard, there is a need for universities to ensure that graduates possess the skills that will enable them to leverage digital technologies and prepare them for the labour market (Jørgensen, 2019).
Hence, research efforts need to provide valid and reliable measures that assess students' needs in advance (Litt, 2013;Ting, 2015).The identification of student's needs will allow higher education institutions to enhance their responsiveness and adaptability and promote programmes for the effective and productive use of digital technologies in the academic environment (Brooks, 2016).

Theoretical Background
At an academic level, the effective use of the Internet is essential because it is the primary medium that students regularly use to access course materials and other information resources, communicate with their peers and professors and conduct course work (Kennedy & Fox, 2013;Gosper, Malfroy & McKenzie, 2013;Shopova, 2014).Research has shown that young people easily adapt to the use of the Internet at a technical level (e.g.accessing services, connecting to a WiFi network) (Kennedy & Fox, 2013;Ng, 2012).In addition, they are very familiar with the use of internet technologies for communication purposes (Bullen, Morgan, & Qayyum, 2011;Gosper, Malfroy & McKenzie, 2013;Shopova, 2014).However, they seem to experience difficulties using tools and services for online information search and evaluation (Hargittai, Fullerton, Menchen-Trevino, & Thomas 2010;Head, 2013;Neumann, 2016) and content creation (Ng, 2012;Kennedy & Fox, 2013).In this regard, there is a need for higher education institutions to close the internet skills gap (Helsper & Eynon, 2010;Hargittai, 2010) by applying assessment processes that will assist them in forming policies and training resources for undergraduate students (Litt, 2013).
One of the most comprehensive and updated measures for assessing the above skills areas is the Internet Skills Scale (ISS) which is proposed by Van Deursen, Helsper andEynon (2014.2016).The development of ISS is based on a theoretical framework which includes four types of skills: Operational, Information Navigation, Social and Creative.Operational skills refer to the most basic technical skills required to use the Internet, such as using browsers to access web applications.The Information Navigation skills refer to navigating through various websites with different layouts and searching, selecting and evaluating the information on the Internet.Social skills relate to using online communication services, interacting with others and exchanging meaning.Lastly, Creative skills are the skills someone needs to create different types of acceptable quality content (e.g.text, audio, video) and publish it or share it with others on the Internet (Van Deursen, Helsper, & Eynon, 2014. 2016).
The ISS was developed to measure the general population's internet skills, including young people.The study herein aims to theoretically inform the ISS and provide a reliable and updated measure that can assess nationally representative samples and other subpopulations, namely undergraduate students.In this regard, it will allow for better comparisons between demographics that help future researchers to determine how different aspects of internet skills vary among groups (Litt, 2013).

Research Methodology
The study sample included 180 undergraduate students who were attending courses at the University of Cyprus during the academic year 2017-2018.Approximately 30% of the students were males, and 70% were females.Additionally, 58.3% belonged to the 18-20 age group, 33.4% to the 21-23 age group, while a small percentage of 8.3% belonged to the 24-25 age group.Finally, 23.9% were first-year students, 31.1% were in the second year, 26.7% were in the third year, 14.4% were in the fourth year, and the remaining 3.9% were in the firth year of study and above.

Description of the Internet Skills Scale
The original Scale includes 35 proposed Likert-scale items on five skills categories: Operational, Information Navigation, Social, Creative and Mobile.The category "mobile skills" (three items) was excluded because the importance of and distribution of skills in using mobile devices was not in this study's scope.Additionally, Van Deursen, Helsper and Eynon (2014) proposed future studies to include new items, such as "critical skills" which refer to the process of evaluation.For this reason, two "critical skills" items were included which were suited for the study's context (I am confident in selecting search results, I carefully consider the information I find online).

Preparation for the validation process
For validation purposes, the original items of the Scale were translated into Greek from two experts in Technology, who then discussed and resolved the discrepancies.A back-ward translation was followed, and the final version of the translated questionnaire was pilot tested on a small sample.
Prior to analysis, descriptive statistics were performed for each questionnaire items to determine whether there were any floor or ceiling effects in the questionnaire.While most of the items performed well, some of the items had very strong ceiling effects.In particular, ceiling effects were evident in six questions of the operational skills set (I know how to open a new tab in my browser (1), I know where to click to go to a different webpage (2), I know how to open downloaded files ( 6), I know how to download/save a photo I fould online ( 5), I know how to connect to a WIFI network ( 9), I know how to upload files ( 8)).Additionally, the same effects were evident in one question of the social skills set (I know how to remove friends from my contact lists ( 23)).This finding is consistent with literature that suggests that younger generations, due to their familiarity and frequent use of the Internet, have already acquired basic technical knowledge (Ng, 2012;Kennedy & Fox, 2013;Ting, 2015;Ståhl, 2017).Because of these ceiling effects, it was considered more appropriate for these items to be removed from the scale.

Results
In order to validate the construct validity of the constructs measured in the Internet Skills Scale, we performed an Exploratory Factor Analysis (EFA) in the remaining 27 items of the Scale with SPSS 24.
Initially, we examined the suitability of the data for factor analysis.The correlation matrix indicated that there were many correlations with a coefficient greater than 0.30 (and no correlation with a coefficient greater than 0.90).The KMO had an initial value of 0.87 and the Bartlett's test suggested rejecting the null hypothesis at the level of α = .01.Factor analysis resulted in a 27-item scale with five subscales: Information-Navigation, Social, Creative, Operational and Critical.Results from the reliability analysis showed that all five factors had high internal reliability (Cronbach's alpha is 0.90 for the first factor, 0.89 for the second factor, 0.91 for the third factor, 0.84 for the fourth factor and 0.87 for the fifth factor).The use of principal component analysis for factor extraction suggested the presence of five factors with eigenvalues > 1, which together accounted for 70.3% of the variance (38.9%, 13.8%, 7.4%, 6.2% and 4%, respectively).For the best interpretation of the factors, the initial pattern matrix was rotated orthogonally using varimax rotation.
The first factor, Information-Navigation skills, explained 38.9% of the variance and was composed of seven items with factor loadings .536 to .904.It is important to note that most of the highest loading items refer to navigation.The second factor, Social skills, explained 13.8% of the variance and loaded clearly with seven items with factor loadings .648 to .810.Although in the original Scale, the items 25 and 26 loaded higher on the Creative skills factor, in this study they loaded higher on the Social skills factor.The third factor, Creative skills, explained 7.4% of the variance and was composed of five items with factor loadings .766 to .879.The fourth factor, Operational skills, explained 6.2% of the variance and was composed of four items with factor loadings .527 to .849.Lastly, the fifth factor, Critical skills, explained 4% of the variance and was composed of four items with factor loadings .565 to .664.The Critical skills set has formed a fifth factor, although the framework refers to the processes of evaluation as part of the Information Navigation skills.However, it is worth noting that two of the items, 18 and 33 in the Critical skills factor load (more than 0.5) to the Information Navigation skills factor.

Discussion
Internet skills research could benefit from developing measures that show reliability and validity.This study performed an Exploratory Factor Analysis to adapt and validate the Internet Skills Scale to the context of higher education in Cyprus.The results produced four factors that remained relatively consistent with the original Scale.The addition of the "critical skills" items resulted in the composition of a new factor and a new set of skills, as the authors of the original Scale suggested.The results from the descriptive analysis showed that the highest ratings were given to items from the Operational and Social skills sets, all of them referring to technical skills.Some items from the Information-Navigation skills set referring to navigation were rated high, while one item referring to information search was among the ones that were given very low ratings.Similarly, low ratings were given to items from the Creative and Critical skills sets.However, similar percentages in a range of responses showed that individual differences exist in different skills sets.The fact above confirms the heterogeneity of skills possessed by younger generations, and it is essential to be considered by higher education institutions.Future work needs to analyse the Scale's robustness with different academic groups, to improve the available instrument.Also, further research should examine the relationship between Critical and Information-Navigation skills.Moreover, other Operational skills could be explored, such as safe browsing, or Information-Navigation skills, such as avoiding plagiarism when using online information resources, which are essential for academic environments.

Table 1 . Proposed Items for the Internet Skills Scale
Sometimes I end up on websites without knowing how I got there.12.I find it hard to find a website I visited before.13.I find the way in which many websites are designed confusing.14.All the different website layouts make working with the internet difficult for me.15.I get tired when looking for information online.16.I should take a course on finding information online.17.I find it hard to decide what the best keywords are to use for online searches.18.Sometimes I find it hard to verify information I have retrieved.

Table 2 . Mean, Standard Deviation and Reliability Values for each Construct (N=180).
The number of each item, as it appears in Table1.**The Information Navigation items were reversed since they were negatively worded. *